电子学报 ›› 2021, Vol. 49 ›› Issue (12): 2407-2420.DOI: 10.12263/DZXB.20200565

• 学术论文 • 上一篇    下一篇

海上边缘计算云边智能协同服务建模

乐光学1, 戴亚盛1, 杨晓慧1, 杨忠明1, 马柏林1, 刘建华2   

  1. 1.嘉兴学院数理与信息工程学院,浙江 嘉兴 314001
    2.绍兴文理学院计算机科学与工程系,浙江 绍兴 312000
  • 收稿日期:2020-06-15 修回日期:2021-07-19 出版日期:2021-12-25
    • 作者简介:
    • 乐光学 男.1963年11月出生,博士,嘉兴学院教授,研究领域包括边缘计算、多云融合与协同服务、无线Mesh网络和移动云计算.
      戴亚盛 男.1993年8月出生,硕士,主要研究方向为边缘计算多云融合与协同服务.E-mail:asdai0591@163.com
      杨晓慧 女.1996年5月出生,硕士研究生,CCF会员,主要研究方向为边缘计算与协同服务.
      杨忠明 男,1998年8月出生,硕士研究生,主要研究方向为边缘计算与协同服务.
      马柏林 男.1961年3月出生,博士,嘉兴学院教授,研究领域包括小波分析、调和分析、智能计算与数学建模.
      刘建华 男.1978年出生,博士,副教授,IEEE和CCF成员,主要研究方向为分布式计算、无线通信、多媒体网络、无线传感器网络和机器学习.
    • 基金资助:
    • 国家自然科学基金 (U19B2015); 浙江省“鲲鹏行动”计划支持项目

Model of Cloud-Edge Cooperative Service for Maritime Edge Computing

YUE Guang-xue1, DAI Ya-sheng1, YANG Xiao-hui1, YANG Zhong-min1, MA Bo-lin1, LIU Jian-hua2   

  1. 1.School of Mathematics and Information Engineering, Jiaxing University, Jiaxing, Zhejiang 314001, China
    2.Department of Computer Science and Engineering, Shaoxing University, Shaoxing, Zhejiang 312000, China
  • Received:2020-06-15 Revised:2021-07-19 Online:2021-12-25 Published:2021-12-25
    • Supported by:
    • National Natural Science Foundation of China (U19B2015); Kunpeng Action Talents Introduction Project of Zhejiang Province

摘要:

由于受环境、资源、能耗、异构等因素制约,海上无线技术发展明显滞后于陆地.以低开销、自适应和自主融合为约束,提出一种海上边缘计算云边智能协同服务策略模型(Model of Cloud-Edge Cooperative Service Scheme for Maritime Edge Computing,MCECS-MEC).基于边缘计算构建海上云边智能协同服务网络框架,抽象海上边缘计算节点行为特征,建立具有抑制联合作弊的节点信任和推荐量化综合评价模型,根据其综合属性评价将准盟员节点融合聚类到不同的协同服务池,实现分级就近服务;基于协同服务请求的优先级和负载均衡理论,设计协同服务池组建规则和段页式自适应轻量级、自适应过热规避盟员发现算法,以状态机方式描述和分析MCECS-MEC协同服务状态演化.基于Router View公开数据集对MCECS-MEC模型性能进行仿真分析,仿真实验表明,MCECS-MEC相比于AODV(Ad hoc On-Demand Distance Vector Routing)、SR(Stochastic Routing)算法,减少了57.7%和55.04%的冗余传输流量,链路重寻率小于3%,负载率稳定于65%.MCECS-MEC模型能有效降低过载、热区、空洞效应等对网络性能的影响,提高海上边缘计算云边智能协同服务效率和质量.

关键词: 海上边缘计算, 云边智能协同, 信任与推荐, 自适应过热规避, 协同状态演化

Abstract:

Limited by environment, resources, energy consumption, heterogeneity, etc., the development of maritime wireless data network technology is backward. A model of cloud-edge cooperative service scheme for maritime edge computing(MCECS-MEC) is proposed. Based on edge computing, it constructs an intelligent cloud-edge cooperative service network framework in maritime edge computing. The abstracted behavior characteristics of nodes in maritime edge computing are clustered into different cooperative service pools by restraining the joint cheating trust evaluation and recommendation quantitative comprehensive evaluation models. Based on the priority evaluation and the theory of load balancing of cooperative service, the rules of building cooperative service pool and the algorithm of segment page adaptive lightweight and heavy load avoidance are designed to discovery cooperative node. The state evolution of MCECS-MEC cooperative service is described and analyzed by state machine. Based on router view open data set, simulation results show that our algorithm reduces 57.7% and 55.04% redundant transmission traffic compared with ad hoc on-demand distance vector routing(AODV) and stochastic routing(SR). The link retransmission rate is less than 3%, and the load rate is stable at 65%. It can effectively alleviate node overload and hot region, and improve the efficiency and quality of service of cooperative service of maritime edge computing.

Key words: maritime edge computing, intelligent cloud-edge cooperation, trusted recommendation, adaptive over hot avoidance, the state evolution of cooperation

中图分类号: